A low frequency persistent reservoir of a genomic island in a pathogen population ensures island survival and improves pathogen fitness in a susceptible host

The co-evolution of bacterial plant pathogens and their hosts is a complex and dynamic process. Host resistance imposes stress on invading pathogens that can lead to changes in the bacterial genome enabling the pathogen to escape host resistance. We have observed this phenomenon with the plant pathogen Pseudomonas syringae pv. phaseolicola where isolates that have lost the genomic island PPHGI-1 carrying the effector gene avrPphB from its chromosome are infective against previously resistant plant hosts. However, we have never observed island extinction from the pathogen population within a host suggesting the island is maintained. Here we present a mathematical model which predicts different possible fates for the island in the population; one outcome indicated that PPHGI-1 would be maintained at low frequency in the population long term if it confers a fitness benefit. We empirically tested this prediction and determined that PPHGI-1 frequency in the bacterial population drops to a low but consistently detectable level during host resistance. Once PPHGI-1-carrying cells encounter a susceptible host, they rapidly increase in the population in a negative frequency dependent manner. Importantly, our data show that mobile genetic elements can persist within the bacterial population and increase in frequency under favourable conditions